David Demelier <mark...@malikania.fr> wrote:
> On Fri, 2022-05-06 at 00:31 +0200, Omar Polo wrote:
> > I don't use pelican, but at glance there are three things to fix in
> > the
> > proposed diff:
> >
> > - remove REVISION: it's incremented every time a port is changed
> > (security patches or changes to some fields like COMMENT or the
> > description) without changing the port' version. When bumping the
> > version, revision rewinds back to the empty string (by removing
> > the
> > variable)
>
> Thanks for the update!
>
> > - this bit of the diff looks odd:
> >
> > > lib/python${MODPY_VERSION}/site-
> > > packages/pelican/themes/notmyidea/templates/authors.html
> > > lib/python${MODPY_VERSION}/site-
> > > packages/pelican/themes/notmyidea/templates/base.html
> > > +@mode
> > > +lib/python${MODPY_VERSION}/site-
> > > packages/pelican/themes/notmyidea/templates/categories.html
> > > +@mode 644
> > > lib/python${MODPY_VERSION}/site-
> > > packages/pelican/themes/notmyidea/templates/category.html
> > > lib/python${MODPY_VERSION}/site-
> > > packages/pelican/themes/notmyidea/templates/comments.html
>
> Yes, not really sure why update-plist added this mode. Since the files
> are correct:
>
> -rw-r--r-- 1 root bin 4303 Feb 9 14:25 base.html
> -rw-r--r-- 1 root bin 404 Feb 9 14:25 categories.html
> -rw-r--r-- 1 root bin 90 Feb 9 14:25 category.html
>
> > - `make test' needs some love: git is still needed as TEST_DEPENDS
> > and
> > the patch needs to be adjusted. some tests fails, but overall
> > it's
> > not that bad :)
> >
> > % make test
> > ...
> > Ran 263 tests in 25.432s
> >
> > FAILED (failures=8, errors=2, skipped=21)
> > Test failed: <unittest.runner.TextTestResult run=263 errors=2
> > failures=8>
>
> Aside the failing tests I don't see any major problem using the
> software, does this mean we can commit your diff?
i wanted to test the runtime a bit too. i've followed the tutorial and
created some dummy post locally and seems to work fine.
i've also removed the dependency on py-six as is not needed anymore.
(it's stil a dependency of at least py-dateutils)
committed, thanks!
> Regards,
